Twenty-one subjects have been treated in a study of midazolam-induced changes in EEG activity, event related EEG responses, vigilance task performance, and self-rated sedation. Results indicate that the effects of midazolam are similar in abstinent alcoholic and nonalcoholic controls. Power analysis indicated that the absence of significant group differences was not an artifact of Type II error.